How to compute 8 days mean from one year data?

I am using the following matlab code

data_matrix=readtable('C:\matlab\sample.csv');
%disp(data_matrix)
[ad,~,cd] = unique(data_matrix,'rows');
out_day = [ad,accumarray(cd,data_matrix,[],@nanmean)];
% Write output as matrix in csv format
filename = 'c:/matlab/sample_2014.csv'
writematrix(out_day,filename);

and my input file is attached as follows

This code is giving error and not computing 8 days average mean.

Please suggest me how to fix it and get the 8 days average mean. The dimension of input file is (365,9) and the dimension of output should be (46,9)

Thanks a lot for your help.

Devendra

Answer 1

[Adam Danz]

Here is a boxcar average of your 365x9 matrix with an averaging window of 8x1 except for the last bin. Averaging is independent for each column producing a 46x9 matrix.

data_matrix=readmatrix('sample.csv');
size(data_matrix)

ans = 1×2

365 9

windowSize = 8;  % 8x1
start = 1:windowSize:height(data_matrix); 
stop = [start(2:end)-1, height(data_matrix)]; 
nbins = numel(stop); 
dataAverages = nan(nbins, width(data_matrix)); 
for i = 1:nbins
    rows = start(i) : stop(i);
    dataAverages(i,:) = mean(data_matrix(rows, :), 1, 'omitnan');
end
size(dataAverages)

ans = 1×2

46 9

Answer 2

[Satwik Samayamantry]

Hi Devendra, Can you please elaborate the problem statement you are working upon? It's not clear what are you trying to achieve.

[Devendra]

I want to compute the 8days average values for 9 variables from 365 days data. Now I am trying a different matlab code do it but it is also giving error.

data=readtable('C:\matlab\sample.csv');

%disp(data_matrix)

T=table2timetable(data);

TT=retime(T,'8-days mean','mean')

% Write output as matrix in csv format

filename = 'c:/matlab/sample_2014.csv'

writematrix(TT,filename);

The dimension of TT should be (46,9)

I hope it is now clear.

Thanks.

Devendra

Answer 3

[Nathan Hardenberg]

I would use the moving average function (movmean)

data_table = readtable('sample.csv');  % reading as TABLE

data_matrix = table2array(data_table);  % convert to matrix

movAvg = movmean(data_matrix, 8, "omitnan");  % moving average with windowsize 8
idx = (1:8:length(movAvg)) + 3  % get indecies where you want to get datapoints

idx = 1×46

4 12 20 28 36 44 52 60 68 76 84 92 100 108 116 124 132 140 148 156 164 172 180 188 196 204 212 220 228 236

eightDayAverage = movAvg(idx, :)  % get the 46 datapoints accordingly

eightDayAverage = 46×9

0.3548 0.3784 0.5569 0.7060 0.4252 0.4261 0.3547 0.4538 0.5223 0.2562 0.1745 0.5648 0.6264 0.3600 0.3550 0.4437 0.2553 0.6969 0.4398 0.2836 0.6227 0.7054 0.3693 0.3653 0.2884 0.2312 0.7599 0.7363 0.5543 0.7236 0.7493 0.3609 0.3551 0.1303 0.3299 0.7261 0.8566 0.8385 0.6312 0.6574 0.4194 0.4178 0.1812 0.6236 0.4623 0.6179 0.7404 0.6243 0.7155 0.3347 0.3425 0.1551 0.7185 0.3177 0.3560 0.5496 0.6344 0.7289 0.4693 0.4755 0.1498 0.7014 0.2785 0.1275 0.1984 0.4755 0.5150 0.5141 0.5080 0.1481 0.4182 0.5088 0.3164 0.2182 0.4501 0.5799 0.5168 0.5134 0.2561 0.2589 0.7062 0.5180 0.3383 0.6504 0.7485 0.3273 0.3230 0.3196 0.2329 0.7749

As an advice: it is always good to paste the full error message into your question. Then it is easier to see what's wrong. In your case you did input the table and not a matrix into the accumarray-function. But fixing this did not work either. Maybe someone else is willing to try it with the accumarray-function.

Answer 4

[Torsten]

data_matrix = readtable('sample.csv');
data_matrix = table2array(data_matrix);
n = floor(365/8);
out_day = zeros(n+1,9);
for i = 1:n
  out_day(i,:) = nanmean(data_matrix(1+(i-1)*8:i*8,:));
end
out_day(end,:) = nanmean(data_matrix(1+n*8:end,:));
out_day

out_day = 46×9

0.3369 0.3542 0.5598 0.6884 0.4191 0.4187 0.3497 0.4362 0.5358 0.2697 0.1771 0.5582 0.6394 0.3646 0.3599 0.4559 0.2455 0.7094 0.4653 0.3048 0.6491 0.7203 0.3585 0.3548 0.2644 0.2368 0.7599 0.7818 0.6129 0.7220 0.7309 0.3712 0.3657 0.1185 0.3699 0.6964 0.8300 0.8237 0.6206 0.6511 0.4006 0.3996 0.1855 0.6284 0.4519 0.5907 0.7275 0.6364 0.7410 0.3470 0.3549 0.1425 0.7266 0.3039 0.3128 0.5161 0.6099 0.7081 0.4876 0.4931 0.1548 0.6957 0.2749 0.1526 0.1837 0.4502 0.4968 0.5344 0.5279 0.1607 0.3718 0.5599 0.3311 0.2219 0.4811 0.5990 0.4775 0.4741 0.2609 0.2497 0.7184 0.5495 0.3639 0.6476 0.7560 0.3265 0.3223 0.3297 0.2389 0.7758